Multiple Input Multiple Output (MIMO) is a telecommunication technology that uses multiple antennas or an antenna array at the transmitter and receiver for signal transmission. The communication channel capacity increases linearly with the number of antennas. The MIMO technology may increase the frequency utilization of the system and expand the system capacity. To improve the system performance, the MIMO system may use the channel information to pre-code signals that are to be sent. To reduce the feedback information, a codebook that is known to the transmitter and receiver can be pre-designed. The codebook includes a series of pre-coding matrixes. The receiver may select a pre-coding matrix from the codebook based on certain rules, and feed back the index of the pre-coding matrix to the transmitter. The transmitter uses the pre-coding matrix to pre-code the information that is to be sent. The process of selecting a pre-coding matrix from a codebook is often referred to as Precoding Matrix Indicator (PMI) selection. During PMI selection, the Signal to Noise Ratio (SNR) at each layer in a system need be calculated, and a pre-coding matrix that results in a high SNR in the system need be selected. The layer here refers to the efficiency (that is, multiplexing rate) of data transmission on the space at a same time point over a same frequency band. If the number of layers in the system is n, it means that the system can transmit n data streams at a same time point over a same frequency band.
According to the conventional MIMO pre-coding technologies, two methods for selecting a PMI are available.
Method 1: For each pre-coding matrix, the SNR at each layer in the system is calculated and the minimum SNR among the SNRs of all layers in the system is recorded. The minimum SNRs of all the pre-coding matrixes are compared. The pre-coding matrix with the highest minimum SNR is selected.
Method 2: For each pre-coding matrix, the average SNR at each layer in the system is calculated. The pre-coding matrix with the highest average SNR is selected.
Shortcomings of the existing technologies: When the pre-coded modulation and coding mode is fixed, the pre-coding technology realized by using the preceding matrix selection methods cannot meet the requirement of a system for different SNRs. The pre-coding matrix selected by using method 1 ensures that the minimum SNR at each layer of the system is the highest, but cannot ensure that the average SNR of the system is high, especially in a system with low SNRs. The pre-coding matrix selected by using method 2 ensures the highest average SNR in the system, but cannot prevent the impact of the lowest SNR at all the layers of the system. When the SNRs in the system are high, the matrix selected by using method 2 may reduce the SNR at the layer with the lowest SNR in the system, thus lowering the system performance.